88 M. De Gasparin on the Classification of Soils. 



being, is much moi'e natural in practice, than that which would 

 consist in considering in granite only the three minerals which 

 compose it, without regarding their aggregation ; and still 

 more than that which, decomposing these minerals into their 

 last chemical elements, would remove granite from nuneralogy, 

 and view it only as a compound of oxygen, silicium, aluminium, 

 potassium, magnesium, and iron. It is thus too with soils, al- 

 though some may present only a single mineral element, as 

 for example, silex ; and though others, as much oftener hap- 

 pens, contain many, and these associated with vegetable and 

 animal debris. We can consider abstractedly each of these 

 mixtures as a pulverized rock, and deal Avith it, as we do with 

 rocks, in forming a systematic whole. 



After having thus demonstrated that both reason and cus- 

 tom authorize us to propose a classification of soils with a spe- 

 cial relation to agi'iculture, we may examine ; 1st, what the 

 characters are which agriculture should examine in soils ; 2d5 

 the relative value of each of these characters ; and, 3d, their 

 application to classification. 



§ I. The Characters of Soils in relation to Agriculture. 



When an agriculturist devotes himself to the investigation 

 of a soil, it is a matter of indifference to him whether it is 

 composed of alumina and silica, or whether these substances are 

 in the state of quartz or felspar, or that by their aggregation 

 they form the debris of granite, or finally, that they belong to 

 primitive, transition, or alluvial formations : what he re- 

 quires, is to know what kind of plants the soil will produce 

 with the greatest advantage, the trouble it will require to 

 put it in a state of culture, the manuring it will need, the 

 quantity of this manure it will yield to the plant, and the por- 

 tion it will retain in its OAvn substance ; these are its agricul- 

 tural characters, — those which adapt it to the objects of agro- 

 nomy, and Avhich shed Hght on his reseai'ches. 



What we have already said of the composition and pro- 

 perties of soils, demonstrates that certain of their scientific 

 elements have a relation to the properties which are inquired 

 after by agriculture. Thus, as to the nature of the cro]>.^ 

 which may be expected from different soils, those which con- 



